It is a difficult task for an unmanned ground vehicle (UGV) to sense obstacles in out fields or unstructured environments. Because the height information is a vital feature to boost the performance of obstacle discrimination, the three-dimensional imaging technique for sensing obstacles ahead UGV of interferometric synthetic aperture radar (InSAR) was presented. The basic signal process flow of InSAR was reviewed. Special factors of the UGV platform that impact the digital elevation model (DEM) measurement precision were analyzed, such as the baseline length, platform motion errors. The DEM of a partial sight-blocked obstacle scene was obtained by processing the three-dimensional InSAR image, which proved the feasibility of applying InSAR to obstacle sensing of UGV.